Improved high-quality MPEG-2/4 advanced audio coding encoder

We propose three methods for implementing an improved MPEG-2/4 Advanced Audio Coding (AAC) encoder. The naive implementation using only the informative part of the AAC standard results in coded audio signals with a significantly poor quality. However, the encoder implementation can be modified to improve coding efficiency. Some coding modules require more sophisticated methods for improving the quality. The first among the three methods is a grouping method in the short block mode of AAC. The second is a codebook selection method in the Huffman coding. The third is the scale-factor sectioning method in bit allocation to the frequency components. We also present the methods proposed in our previous papers. The efficiency of the proposed methods has been evaluated objectively and subjectively. We have confirmed that the proposed methods are capable of improving the quality of encoded streams.